Study suggests bias against 'black' names on resumes

HR Magazine, Feb, 2003 by Bill Leonard

Employers may be selecting or overlooking prospective job candidates for interviews based on their potential race as suggested by names, according to a recent study by two professors from the University of Chicago and Massachusetts Institute of Technology (MIT).

To test whether employers might discriminate against job applicants with black-sounding names, associate professors of economics Marianne Bertrand with Chicago's Graduate School of Business and Sendhil Mullainathan with MIT conducted an elaborate experiment. They fabricated resumes for multiple "phantom" job seekers with common black and white names. The professors then sent out nearly 5,000 resumes for 1,300 job openings advertised in newspapers and on online job sites throughout Chicago and Boston.

"We searched online and selected resumes of actual job seekers," says Bertrand. "We then used those to create models for several different realistic resumes with the appropriate education and experience needed for typical job openings advertised in newspapers."

Most job openings for which the researchers sent resumes were administrative, sales, clerical and managerial positions. Bertrand and Mullainathan randomly assigned the applicants names common to either black men, black women, white men or white women and were careful not to send identical resumes to the same employer.

Bertrand and Mullainathan then tracked which of their "applicants" were called for job interviews. Bertrand said that more resumes were sent to Chicago area employers simply because it is the larger metropolitan area but added that the rate for interview requests was virtually identical between the two cities.

The results are a bit disturbing, the researchers admit. Applicants with white-sounding names were 50 percent more likely to be contacted for job interviews than those with typical black names. There were no significant differences between the rates at which men and women were contacted.

"Once we compiled the data and got a good look at it, I was immediately struck by the disparities in the response rates," said Bertrand. "I expected that there would be a difference, but not one that was so striking."

Bertrand acknowledged that employers could not conclusively determine the race of a job applicant from a resume. However, she contended that the study clearly showed a bias among employers based on perceptions.

"I believe our study clearly shows that when employers or hiring managers see a name that might indicate race, or at least the applicant's social class, they appear to react negatively to-wards it," Bertrand says.

"Once you take into account the large size of our sample and the comparable response rates of the two cities, we believe it is a very good indicator that employers treat applicants with African American-sounding names much differently Bertrand says.